Carry Trade — Profiting From Rate Differentials (and Why It Blows Up)

Most trading strategies try to predict where price will go. Carry trade does something else: it profits from the fact that money is priced differently across currencies. You borrow the currency of a low-interest-rate country, buy the currency of a high-interest-rate country with it — and collect the interest difference every day. The exchange rate can sit still, and the account still grows.

For decades this was one of the pillars of global speculation — with the Japanese yen playing the role of eternal funding currency. And it was the yen that showed carry's other face in August 2024: when carry unwinds, it doesn't walk down the stairs, it jumps out the window. This article covers the mechanics, the arithmetic, and — honestly — the anatomy of the blow-up.

Educational disclaimer: this material is for educational purposes only and is not investment advice. Carry trade on the currency market uses leveraged instruments (CFDs, futures contracts) — instruments with a high risk of rapid losses; a significant majority of retail CFD accounts lose money. Historical rate differentials do not guarantee future flows. Don't commit funds you can't afford to lose.

How Carry Trade Works

Every currency has its own interest rate, set by a central bank. When you hold a currency-pair position overnight, the broker settles the difference in interest between the two currencies — the so-called swap or rollover points. Are you long the higher-yielding currency and short the lower-yielding one? Interest flows to you. The other way around — you pay.

Carry trade means positioning yourself on the receiving side of that flow:

1. Funding currency — you borrow (go short) a currency with low rates. Classically: the Japanese yen, and for years the Swiss franc too.
2. Target currency — you buy a currency with high rates: historically the Australian or New Zealand dollar, in the 2022–2024 cycle mainly the US dollar, and in the aggressive version the Mexican peso or Turkish lira.
3. Flow — a daily positive swap, proportional to the rate difference and the position's notional.

In practice you don't have to manually borrow anything — a long USD/JPY position with a broker is technically exactly this construction: long dollar, short yen, swap settled automatically.

The same logic shows up in crypto under a different name: the funding rate strategy on perpetuals also earns on a payment flow instead of on direction. The difference: classic FX carry is usually not currency-hedged — and it's precisely the exchange rate that's the main risk.

What the Numbers Say

Textbook arithmetic (the 2023–2024 cycle). Fed rate: 5.25–5.50%. Bank of Japan rate: −0.1% to 0.25%. Difference: roughly 5 percentage points a year in the dollar's favor.

That 50% is a bait number, and it needs to be defused right away. First, the broker won't hand you the full rate difference — retail swap points tend to run 1–2 p.p. worse than theory. Second, and more importantly: 10x leverage means a 10% move against the position wipes out the deposit. And that's exactly the kind of move that hit USD/JPY within a single month of 2024.

Case study: the yen unwind, August 2024. In early July 2024, USD/JPY was trading around 161–162 — the weakest yen in nearly 40 years, with speculative positioning short the yen at a record one-sided extreme. Then three dominoes fell: on July 31 the Bank of Japan raised its rate to 0.25% and signaled further tightening; on August 2, weak US jobs data raised expectations of Fed rate cuts (the differential narrowing from both sides at once); leveraged traders started closing yen shorts, which strengthened the yen further and forced more closures.

The result: by August 5, 2024, USD/JPY had fallen to around 141–142 — roughly a 12% strengthening of the yen in a little over a month — Japan's Nikkei posted its worst percentage session since the 1987 crash (about −12% in a single day), and the sell-off spilled into stocks and crypto globally. Investment bank analyses at the time estimated that most of the speculative yen carry trade was unwound within a handful of sessions. For a trader long USD/JPY at 10x leverage, that wasn't "a bad month" — it was a wiped-out account, with more than a year's worth of future interest evaporating in three sessions.

This wasn't an anomaly, it was a pattern: carry gains have a distribution of "often small," losses of "rarely, but huge." In market literature, carry trade is often described as picking up coins in front of a steamroller.

Why this works at all (and why theory says it shouldn't). Textbook interest rate parity theory predicts that the higher-yielding currency should weaken by exactly the amount of the rate differential — the interest gain would be eaten by the exchange rate, and carry would pay nothing. Empirically, for decades the opposite often happened: high-interest-rate currencies on average didn't depreciate the way theory predicted (in the literature this is the so-called forward premium puzzle), so systematic carry across baskets of currencies delivered positive returns. The catch is in the word "on average": that excess return is largely a premium for the risk of rare crashes — exactly like August 2024. You're not getting paid for being clever; you're getting paid for being exposed to a scenario that will eventually arrive. Professionals soften this with a basket: spreading carry across several currency pairs instead of one reduces dependence on a single central bank — though in a global risk-off event, most legs unwind at once anyway.

How It Works Step by Step (A Numerical Example)

An illustrative example, rounded figures:

1. Pair selection. You look for a pair with a clear, stable rate difference and a dovish central bank on the funding side. You check your broker's actual swap points (not the macro-table theory) — e.g., long USD/JPY: +$12 a day per $100,000 notional.
2. Trend filter. Carry makes sense when the exchange rate is at least not falling. Entry condition: rate above its long-term average, no reversal signals from the funding currency's central bank. Entering "because the swap looks nice" without checking the chart is asking for another August 2024.
3. Position size. $20,000 capital, leverage capped at 2–3x (not 10x): a $50,000-notional position. Swap: ~$6 a day, ~$2,200 a year — 11% on capital from the flow alone.
4. Stop and monitoring. A technical stop below the last significant support, e.g., 4% from entry — risking $2,000 (10% of capital? too much — cut the position in half or tighten the stop; run the risk math the same way as in position sizing). Mark every meeting of both central banks on the calendar.
5. Exit. The signal to close isn't "the swap stopped feeling good," it's a change in the fundamentals: the funding-currency central bank starts raising rates, the target-currency central bank starts cutting, or the exchange rate breaks its trend structure. That's when you exit — even if the daily flow is still positive.

Risks — When the Strategy Doesn't Work

• The unwind, i.e., the crowd unwinding. Risk number one. Carry is inherently a herd trade — when everyone sits on the same side, any impulse (a funding-side rate hike, risk-off, weak data) triggers an avalanche of closures. The funding currency strengthens in a jump, and liquidity disappears exactly when you want to exit. Gains by the stairs, losses by the elevator.
• Distribution asymmetry. For months you earn a few dollars a day, then give back a year's flow in a week. A high percentage of winning days creates an illusion of safety — statistically this is a negatively skewed strategy, psychologically the worst possible profile for someone with no exit plan.
• Leverage. The rate differential alone (3–5% a year) is too small to excite a speculator, so the industry adds leverage. But leverage doesn't increase the edge — it only increases the amplitude, and with negative skew, amplitude kills.
• A narrowing rate differential. The flow isn't eternal: the funding-side bank can hike, the target-side bank can cut — the differential can shrink from 5 p.p. to 2 p.p. within a few quarters and stop covering costs and risk.
• Broker costs. Retail swap points are systematically worse than the theoretical rate difference; at some brokers the positive side of the swap is symbolic. Compare actual rates before you calculate a "profit."
• Exotic currencies. The peso, lira and forint tempt with double-digit rates, but a high rate is the price of inflation and devaluation risk — it can give back more in the exchange rate than it pays in interest. That's not a bonus, it's a risk price tag.

Verdict, no hype: carry trade has a real, structural source of profit — the price difference of money — and over long stretches it works quietly and reliably. But it's a strategy with an "insurer's" profile: you collect small premiums, and once every few years you pay out a claim. August 2024 wasn't an accident, it was a bill for years of one-sided positioning. If you trade carry: low leverage, a trend filter, both central banks' calendars on the fridge, and an exit plan written before you enter. Without that, you don't have a strategy — you have a deferred loss with an interest coupon attached.

FAQ

Rafał — Krypto Bez Ściemy

Trader and founder of Strefa Tradingu. He’s been breaking crypto down on YouTube for years — no hype, no signals, with a focus on market structure and risk management.

kryptobezsciemy.plInstagramYouTubeKlub Zero